Your competitors already did some of your keyword research for you. Every page that ranks for them is a public record of a search query that real people type, that Google is willing to send traffic for, and that a site like yours can win. A keyword gap analysis turns that record into a list of topics you are missing.
Done badly, it produces a spreadsheet of thousands of keywords nobody acts on. Done well, it produces a short list of pages worth writing, in the order worth writing them. This guide covers the second version.
What a keyword gap actually is
A keyword gap is a search query where at least one competitor ranks and you do not, or where you rank far below them. There are three kinds, and they call for different work:
- Missing: competitors rank, you have no page on the topic at all. This usually means a new article.
- Weak: you have a page, but it sits on page two or three while competitors are in the top ten. This usually means improving the page you have, not writing another one.
- Shared but losing: you and competitors both rank on page one, but they sit above you. These are often worth a title, intent or internal linking fix before anything bigger.
Mixing these up is the most common mistake. If you write a new article for a weak gap, you end up with two pages competing for the same query, and often neither ranks.
Step 1: Pick the right competitors
The competitors that matter for SEO are not always your business competitors. A company that sells the same product as you but has a small blog tells you little. A publication, a tool directory or an adjacent product that ranks for the questions your customers ask tells you a lot.
A practical way to find them:
- Write down five to ten queries a customer might search before they need you, for example "how to plan a content calendar" rather than your product's name.
- Search each one and note which domains appear on page one more than once.
- Keep three to five domains that show up repeatedly and that are within reach. A site whose pages are similar in depth and authority to yours is a better benchmark than the largest site in your space.
Step 2: Collect the keywords they rank for
Any SEO data tool with a "ranked keywords" or "organic keywords" report will give you the queries a domain ranks for, with an estimated position and search volume. Export the list for each competitor and for your own site.
Keep the filters loose at this stage, but drop the obvious noise:
- Branded queries (a competitor's own name and product names).
- Queries where the competitor ranks below position 20. They are not winning those either.
- Queries with no estimated volume at all, unless they are clearly high intent for your product.
Step 3: Find the gaps
Compare the lists. For each query, record the best competitor position and your own position, then sort the query into one of the three kinds above.
| Query | Best competitor | You | Kind of gap | Likely action |
|---|---|---|---|---|
| content calendar template | #3 | not ranking | Missing | New article |
| how to write a content brief | #5 | #24 | Weak | Improve the existing page |
| seo content checklist | #2 | #7 | Shared but losing | Title, intent and links first |
Group queries that mean the same thing before you count them. "Content brief template", "content brief example" and "how to write a content brief" are usually one page, not three. Grouping turns a list of hundreds of keywords into a few dozen topics.
Step 4: Check what Google actually rewards
A gap is only an opportunity if you can publish the kind of page that ranks. Before you commit to a topic, look at the current first page for its main query and ask:
- What format wins? Guides, lists, templates, tools, product pages or videos. If the first page is all free tools and you plan a blog post, the post will struggle.
- Who ranks? If every result is a household-name domain, the topic may be out of reach for now. If smaller sites rank, so can you.
- What is missing? Outdated examples, no worked example, no template, shallow answers to an obvious follow-up question. This is your reason to exist on that page.
If you cannot name one concrete thing your page will do better than the current results, skip the topic. Another article that says the same thing in a different order rarely earns a place.
Step 5: Prioritise the shortlist
Score each remaining topic on a few simple questions instead of a single volume number:
- Relevance: would someone searching this plausibly become a customer, or at least care about what you do?
- Demand: is there meaningful search volume across the whole group of queries, not just the main one?
- Winnability: do sites like yours rank on page one?
- Effort: is this a 1,200-word guide, or a tool and a dataset?
Weak gaps on pages you already have often come out on top. Improving a page that is already on page two is usually faster than starting a new one from zero, because Google already associates that page with the topic.
Step 6: Turn each topic into a page
For every topic you keep, write a short brief before you write the article:
- The main query and the related queries the page should answer.
- The search intent in one sentence: what the reader is trying to get done.
- The format that matches the first page.
- The one thing your page will do better than the current results.
- The existing pages on your site that should link to it, and the ones it should link to.
Then publish, and give each page a fair window before you judge it. Check impressions and average position in Search Console after four and eight weeks. Pages that gain impressions but few clicks usually need a better title and description. Pages that gain nothing usually missed the intent.
Common mistakes
- Chasing every gap. A competitor ranking for something does not make it right for you. Relevance comes first.
- Ignoring pages you already have. New articles on topics you already cover split your rankings instead of adding to them.
- Copying the competitor's page. Matching the format is good. Matching the content gives Google no reason to rank you above the original.
- Doing it once. Competitors publish, rankings move and new queries appear. A gap analysis every quarter keeps the list honest.
Doing this without the spreadsheet
The steps above work by hand, and doing them once is the best way to understand your market. They are also repetitive: collect rankings, compare, group, check the results page, score, brief, repeat every few months.
Growcake runs this loop for you. It reads your site, finds the queries your competitors win that you do not, checks each one against the live results, and tells you whether the topic needs a new article or an improvement to a page you already have. You see the evidence behind every recommendation, and nothing is written unless research shows a real way to beat what already ranks.